There is compelling evidence indicating that the presence in tumor cells of the DNA repair protein, omicron6-alkylguanine-DNA alkyltransferase (AGT), imparts resistance to chemotherapeutic chloroethylating and methylating agents. Previous studies have shown that omicron6- benzylguanine (BG) inactivates human AGT and that treatment of tumor cells with BG sensitizes them to killing by drugs such as BCNU and temozolomide. Depletion of AGT by BG also improves the therapeutic index for the treatment of human tumor xenografts in nude mice by BCNU and clinical trials of this combination have just begun. The aims of the present experiments are: (a) to identify other AGT inhibitors that would be improvements over BG with respect to potency, ease of delivery, specificity towards tumors and the ability to inactivate AGTs resistant to BG and (b) to provide a greater understanding of the mechanism(s) responsible for the BG resistance of AGTs. The detailed specific aims are: (1) to test compounds designed to be improved inhibitors for the ability to inactivate pure human AGT and a BG-resistant mutant and to modulate AGT activity in cultured tumor cells. The effects of such modulation on killing by BCNU will also be examined; (2) to study the interaction of BG and other known inactivators of AGT with the protein. For this purpose, kinetic measurements of the interaction of the AGT and the inhibitors will be made and the ability of the compounds to act as inhibitors of the formation of [8-3H]guanine from [8-3H]BG will be used as an assay; (3) to study the structure of AGT (and the related Ada-C alkyltransferase from E. coli) and mutants with altered reactivity for BG and other inhibitors. The crystal structure of the AGT protein in the presence and absence of inhibitors and a DNA substrate will be determined. Some of these studies will use an inactive C145A mutant AGT which binds substrates but cannot react with them to form the S-alkylcysteine at the active site; (4) to investigate the spectrum of mutants in human AGT that can lead to resistance to BG. A system for the general identification of such mutants will be used to determine the range of sites at which such alterations can occur by expressing the human AGT in E. coli thus conferring protection from MNNG or BCNU. Such protection will be abolished by BG and after mutagenesis of the plasmid containing the human AGT cDNA, variants which can provide protection to alkylating agents in the presence of BG will be isolated, sequenced and AGT protein prepared for comparison with control AGT for inactivation by BG and the other inactivators.